Did Michael Jackson Abuse Because He Was Abused?

The science on the cycle of abuse is much more complicated than you might think.

Michael Jackson and his father, Joe Jackson, at the Golden Globes.
Michael Jackson and his father, Joe Jackson, pose at the 1973 Golden Globe Awards. Frank Edwards/Fotos International/Getty Images

When Amanda Robson’s husband, Wade, first told her that he had suffered years of childhood sexual abuse at the hands of Michael Jackson, she grew worried for the safety of their son. “Is there any confusion as to what is appropriate in how you love [him]?” she says she asked Wade, in the HBO documentary Leaving Neverland.

Her concern, that as an alleged victim of molestation her husband may have been a threat that spread that evil to another generation, reflects a view so embedded in the culture that it rarely merits much discussion. But the “cycle of abuse,” as it’s often understood, is nothing less than psychological Lysenkoism: the belief, in the absence of clear, empirical evidence, that lived abuse tends to propagate across generations and reproduce itself in specific ways. The facts about the “intergenerational transmission of maltreatment,” as scholars call it, are very complicated, and while there’s some truth to the idea as it’s generally portrayed, there’s also a fair amount of falsehood.

Michael Jackson’s story, in particular, has long been framed by this assumption. His most eccentric habits—the private amusement park and movie theater, the slumber parties that he often had with prepubescent boys, and so on—were often said to be the outcome of the performer’s own troubled and pilfered childhood. Jackson was so afraid of being in the presence of his physically abusive father, Joe Jackson, that it made him vomit, he told Oprah Winfrey in a 1993 interview. His sister Latoya Jackson also said that Joe beat Michael and alleged that he sexually abused both her and another sister too. (There were unsubstantiated rumors that Michael was also a victim of sexual abuse.) This history has for years been cited as if it were an explanation: “All of Jackson’s oddities seem to be reactions to what he suffered as a child,” wrote Jacob Weisberg in Slate in 2005, in what was then, and still remains, a conventional appraisal of the facts.

The same pop psychology was applied again just last month, in another TV documentary about a superstar musician and his alleged sexual abuse of children. Lifetime’s Surviving R. Kelly includes footage of the singer’s 2012 interview with Tavis Smiley, in which Kelly reports that he was himself molested by someone in his family starting at the age of 7. “When you’re a child, whatever you’re taught, it goes with you,” Kelly says. A clinical psychologist, who appears on-screen shortly after, concurs. As victims of such abuse get older, she suggests, they “want to make sure that they’re always in that power position … and there’s really no more powerful position in a sexual relationship than to be the abuser of a child.”

The folk wisdom undergirding this idea might be older, but the theory of intergenerational transmission of abuse, as it’s propounded by criminologists and psychologists, seems to have taken hold in the middle of last century. In 1940, two psychiatrists, Lauretta Bender and Frank Curran, published observations of homicidal children and adolescents who were under their care in New York City. They concluded that many identified themselves with “aggressive parents” and seemed to “pattern after their behavior.”

In the early 1960s, this suggestion that the abused become abusive to others was taken up again amid a burgeoning concern over the physical battery of children. It seemed as though instances of child abuse were increasing—“Cry Rises From Beaten Babies,” warned a Life magazine story in 1963—and clinicians wondered whether victims might internalize and re-enact the violence. “One possible consequence which is overt, obvious and of great public concern,” wrote clinician George Curtis in the American Journal of Psychiatry that year, is “the probable tendency of children so treated to become tomorrow’s murderers and perpetrators of other crimes of violence.” Children might be learning from their parents’ uncontrolled hostility, he said: “The vernacular idea, ‘monkey see, monkey do,’ expresses the same idea in terms understandable to everyone.”

What began as a series of clinical observations and case histories would be elaborated, in the 1970s, with more careful, cross-sectional research. Researchers began to survey groups of juvenile delinquents to see how many of them reported having been abused—and the proportions turned out to be startlingly high. A more recent audit of adult men incarcerated at a county jail (most of whom had been charged with felony crimes), found that 59 percent of the inmates said they’d been sexually molested by the age of 13.

These studies seemed to show a sturdy link between those who were victims of abuse and those who perpetrated crimes. But there were major limitations to this research. For one thing, responses to such surveys can be unreliable or self-serving. Given widespread belief in the “cycle of abuse,” for example, certain inmates might report trauma as a means of excusing their transgressions. Then there’s the problem of causality: It was always hard to tell whether childhood abuse on its own really spurred the later crimes; it could just as well have been the case that other common factors and experiences—such as being from a broken or impoverished home, or having parents with substance abuse problems or mental illness—put individuals at a higher risk of both being victimized in childhood and committing crimes.

The science gets still more dizzying when you try to focus on the “cycle of abuse” as it’s most often understood in pop culture—and the version cited in both the Michael Jackson and R. Kelly documentaries. In this formulation, those with troubled childhoods don’t just go on to act in anti-social ways. It posits, rather, that people do exactly unto others as was done unto them, i.e., that former victims of childhood sexual abuse are uniquely inclined to end up as the perpetrators of childhood sexual abuse. Since different forms of child maltreatment often go together—a single child may be victimized in multiple ways—as do different types of criminal behavior, this only leads us further into thickets of confounding and confusion.

Psychologist and criminologist Cathy Spatz Widom was the first to make some progress through the bramble. In 1989, she published data on the cycle of abuse with a novel methodology. Instead of looking retrospectively at criminals and delinquents, she started by picking out a group of victims of abuse, then following up throughout their lives to figure out what happened. She began her work by identifying more than 900 victims of abuse and neglect whose cases had been registered in the court system of an unnamed Midwestern city between 1967 and 1971. Then she set up a control group, matching up those victims as best as she could with people of the same age, race, and sex who attended the same schools and lived in the same neighborhoods. Finally, she pulled any official records of their delinquency, detention, or adult criminal activity across the next 20 years.

Using this much more powerful and better-controlled design, Widom was able to confirm that victims of childhood abuse are indeed at greater risk of becoming criminals. Perhaps more importantly, she showed that mere neglect—even in the absence of any violent physical abuse—was a noteworthy predictor of later criminal behavior.

She kept following her subjects, who are now well into middle age, and also gathered information from their children. In 2015, Widom published several decades’ worth of further data. One of her papers in particular focused on the question of whether someone’s experience of childhood abuse can predict their sexual offending later on. While 4.5 percent of the people in the control group had been arrested for a sex crime, nearly twice as many—8.3 percent—of the people who had been victims of abuse or neglect went on to perpetrate such a crime. So there was a link, but the details didn’t fit the expected pattern of “monkey see, monkey do.” The people in Widom’s study who were abused as children in specifically sexual ways did not, in fact, appear more likely to get arrested for a sex crime later on; instead, it was the ones who were either neglected or physically abused who ended up at higher risk.

That may have been a quirk of Widom’s data set. Among both groups who had been arrested for a sex crime, almost all of them—84 percent—were men. Yet her study included just two dozen male victims of childhood sexual abuse, of whom three went on to be sexual offenders. It may be that this sample was too small for a true effect to show up in her statistical tests.

A similar study, published in 2016, looked at records of childhood sexual abuse and sexual offending in a group of more than 38,000 Australian men. Among those who had been molested, just 3 percent went on to commit a sexual offense. That rate was much higher than what was found among the total population (0.8 percent), suggesting a cycle of abuse. But being victimized by other forms of childhood mistreatment was also associated with committing sexual crimes, and there were no clear signs of a special one-to-one relationship in which sexually molested children grew up to be sexual molesters.

But really—it’s complicated. A paper published two weeks ago combined and analyzed findings from 142 different studies of intergenerational transmission of maltreatment. The study’s authors, led by the University of Calgary’s Sheri Madigan, concluded that there is indeed evidence for a “modest association” between someone suffering abuse and then perpetrating it, and that specific forms of abuse may be passed down in this way.

Another paper, also published two weeks ago, was among the first to look at whether the tendency to abuse a child might be genetic—perhaps the children of abusive parents are more likely to end up as abusers because they’ve inherited an innate proclivity. That study found evidence in favor of this notion for emotional, but not physical, abuse.

Even if it’s true that victims of childhood abuse are at greater relative risk of committing crimes, including sexual assault, the most important numbers are the ones in the background. The absolute risk that any given victim of sexual abuse will go on to become a sexual offender is very, very low. The Australian study, for example, found that among male sexual offenders, 96 percent had no official history of having been sexually abused. More than four-fifths had no official history of being abused in any way.

In light of those statistics, it’s worth asking why we give so much credence to the “cycle of abuse” as an explanation. Maybe it feels safer to assume that cruelty is predictable, and that something so disturbing as the mistreatment of a child could be circumscribed and contained, drawn into and kept inside a pattern. But the science is filled with too many nuances and caveats to allow such a clear-cut explanation.

At least we know this much: While Michael Jackson and R. Kelly reportedly may have been a bit more prone to be arrested for mistreating children, there is no trodden pathway from victim to abuser. The cycles often stop.